Saved in:
Bibliographic Details
Main Authors: Khilchuk, Maria, Markov, Ilya, Hvatov, Alexander
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.12666
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909962193600512
author Khilchuk, Maria
Markov, Ilya
Hvatov, Alexander
author_facet Khilchuk, Maria
Markov, Ilya
Hvatov, Alexander
contents In differential equation discovery algorithms, numerical differentiation is usually a fixed preliminary step. Current methods improve robustness with data subsampling and sparsity but often ignore the variability from the differentiation method itself. We show that this choice systematically introduces uncertainty, affecting both equation form and parameter estimates. Our study indicates that high-resolution schemes can magnify measurement noise, while heavily regularized methods may mask real physical variations, which leads to method-dependent findings. By evaluating six differentiation techniques on various partial differential equations under diverse noise levels using SINDy and EPDE frameworks, we consistently notice methodological biases in the determined models. This underscores the importance of selecting differentiation methods as a key modeling choice and highlights a path to enhance ensemble-based discovery by diversifying methodologies.
format Preprint
id arxiv_https___arxiv_org_abs_2512_12666
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Differentiation methods as a systematic uncertainty source in equation discovery
Khilchuk, Maria
Markov, Ilya
Hvatov, Alexander
Symbolic Computation
In differential equation discovery algorithms, numerical differentiation is usually a fixed preliminary step. Current methods improve robustness with data subsampling and sparsity but often ignore the variability from the differentiation method itself. We show that this choice systematically introduces uncertainty, affecting both equation form and parameter estimates. Our study indicates that high-resolution schemes can magnify measurement noise, while heavily regularized methods may mask real physical variations, which leads to method-dependent findings. By evaluating six differentiation techniques on various partial differential equations under diverse noise levels using SINDy and EPDE frameworks, we consistently notice methodological biases in the determined models. This underscores the importance of selecting differentiation methods as a key modeling choice and highlights a path to enhance ensemble-based discovery by diversifying methodologies.
title Differentiation methods as a systematic uncertainty source in equation discovery
topic Symbolic Computation
url https://arxiv.org/abs/2512.12666